Fixed tests did not confirm a positive effect from earlier target gaps in nine EEG and thirteen fMRI adults

We tested a positive response coefficient for the interval before the last auditory target. We controlled the exact event count and physical time since the last target, response time, and run trends. Separate public studies supplied scalp EEG and BOLD fMRI. Development used three EEG adults and four MRI adults. A local code lock preceded tests in ten other EEG adults and thirteen other MRI adults. One EEG adult in each stage failed fixed eligibility rules. The reserved fits used 1,423 EEG targets and 593 MRI targets. Mean coefficients were 0.019718 response SD and -0.001773 peak BOLD percent per residual history SD. Their 95% person Student intervals were [-0.032990, 0.072426] and [-0.031652, 0.028106]. Positive counts were 6/9 and 7/13. The fixed shared rule for mean and majority failed, p=0.550355. No positive cross-modal memory effect was confirmed. All negative people and fixed checks remain. Added unfitted examples test response carryover without an older-history neural amplitude. They do not estimate human bias. We matched all 90 original event tables to the compact inputs. Independent calculations checked all eligible coefficients and one original neural run per modality. The package gives code, compact inputs, 2,016 selected target rows, source pins and terms, all outcomes, figures, and repeat steps. Small cohorts, response selection, model form, filters, response overlap, region placement, and distinct tasks limit scope. The result does not establish absent memory, a shared source, or causality.

EEG · fMRI · auditory oddball · event history · P300 · response overlap · reproducibility